Right back from our Scandinavia tour! Shot a lot of these fantastic landscapes at the west coast of Sweden and than also some in the North of Denmark. The mixed weather was kind of cool for doing all sorts of things photographically – from under-water towards copter night time and sunset shots and also using slomos at the coast. I am also feeling much more confident now that the Phantom copter systems are dealing well with strong winds ;).
A shot from Hirtshals Fyr: – flying at night is likely the most exciting you can do with a copter.
Did also some footage with grading in Adobe AE towards nice slomos. I am still working on these but there will be a slomo teaser soon for this project. Its just a very cool feature that can be used nicely to visualize waves and coastal textures and I want to use the 12o fps in the future also for the sailing sport photography. This is clearly a domain where we can do much more – and better ;).
Right now: preparing for the German Classics 2017 – this is as usual one of these very big events. The forecast is kind of hinting windy conditions but this is pretty much expected this summer.
mtk – Sören
Some of the new interesting perspectives of copter based regatta and yacht photography clearly come with time series visuals – or as I would rather call it: “motion-time stacks”. Its not new in sports photography and with fast cont. autofocus and high speed cont. shooting speed we have seen it in motorbike sports, skiing , swimming and especially with athletics and dancing. The appealing aspect is that you can tell a (very short) story with these stacks. The image reveals a series of decisions and is much more than a capture of a specific situation.
With UAV visuals you can create comparable shots – mainly possible due to the precise GPS controlled position hold of the modern quadrocopters – especially the DJI Phantom series. The lat/lon/altitude position is very precisely corrected and together with the gimbal correction you have only minimal changes of the area covered with the FOV of your camera. Combining different situations at the same position is more a question of how to post process the time series and how to combine the changes into one image. With the new Phantom copter generations you can easily leave the copter at a specific place & altitude and let it capture the scene while shooting with a DSLR at the water level. The only problem are winds above 5Bf and/or gusts that exceed 5-6 Bf. The copter drifts away and you would have to manually move the copter in S-Mode to get back to your start position. Flying freely without VR-googles is also not too much advised when you do not have constant visual control of the copter position. You easily loose track where the copter exactly is. The standard RTH function does not work on water (you usually drift on a small rib, when you trigger RTH the copter will fly towards the position where you have been – not to where you are now) – though flying back to sender (and not to the start position) is helpful and implemented in a way as you can update the “Home Point” manually in the Phantom series control software when you have a GPS enabled controller device (iphone/iPad). This can safe your copter because on water you easily loose track where exactly your copter is. RTH with the hover option enabled (you do not want the copter to land automatically!) is a great function as long as you update the home point regularly on the boat. Thats the function that I trained a couple of times for these not so nice moments when you believe that you lost control and the battery warning noise is slowly eating your coolness. Btw: updating the controller position should be set with a shortcut once selected – its the function that you want to trigger every 20sec or so on a boat. I dont understand why this is not already implemented in the DJI GO app.
The „Dynamic Home Point„ feature of the Inspire copter controller software (version update 29.05.17) is also a very remarkable feature in that context. It updates the position of the „Home Point“ using the controller GPS information. The RTH function is in that form a really useful function again. This is by far the most secure option for flights over water where you start from a boat but it doesnt create safe automatic landings on a moving boat! Landing is usually by picking the copter in air by hand. Its the reason why the Phantom series works well here: you can easily pick the copter up by its legs.
Some picture examples that I did in may/june 2017 to find out what is working are attached here. Imo you either fly very high (you need a permission for above 100m!) or very low – approx. 20m – to get exciting perspectives – the in between wont make it in my opinion. Low flying is difficult, you often cannot clearly define the relative height difference to the boat rigs and you do not want to distract peoples attention. Very high positions are nice to capture the field at the start leg or at the luv position but even with the 20mm (equiv) of the P3 you will need 150-250m flight altitude – and clearly a BFA permission to go that high.
To create appealing visual compositions you have to pre-visualize the direction and possible crossings of the boats – that is really the most challenging part. I usually try to get a fixed position close to a turning point where maneuvers take place. There are endless option how to combine light and action into something appealing. With the wide angle 20mm equiv. camera system of the P3 one always collects sun glint that will mess up the stacking process later. Its hard to avoid but you can try to keep at least the action out of the glint region and it can be also nicely used as part of the composition.
With the new copter regulation in place for Germany its easier to do commercial/research work than it was before. Mainly because you do not need an allowance for every flight from the regional authority anymore. There is a general clearance for commercial flights below 100m, and below 5kg weight in place. Above 100m and in EDR zones and some other restricted region-types a special permission is needed (an insurance is also a given).
In Berlin the EDR4 is blocking the south-western part of the Havel – so everything west of the “Kälberwerder” island is very clearly within the EDR4 zone (a restricted flight zone around the HMI experimental nuclear reactor). For this year for the “Havel Klassik” regatta race I applied for a clearance (that I received already) because the interesting visuals are all west of the Peacock Island and fully within the EDR4 zone.
I will add some more shots here in the future to make this post a diary of the experiences with the P3A and the P4P from DJI. So stay tuned.
Have really fully settled on the OSX operating system and Apple hardware and in the last years I was indeed very satisfied – not necessary in terms of performance but clearly ergonomics and usability of the Apple ecosystem is working and most important it just lets you concentrate on doing things. Thats clearly what I learned to appreciate – and yes I also was a Linux fan in the early days and for mass satellite image data processing I am still using Linux servers as number crunchers but for everything else / DTP/Imaging the Mac is it – may be “was it”? The Late2016 version of the trusty workhorse the MacBook Pro is just a very mixed bag. Performance wise the fast SSD is nice and faster GraCa is good – so yes its faster but not by very much! Agisoft Photoscan is clearly very much accelerated – I guess by factor 2 (compared to the full speced-out 2013 model) but thats mainly due to the faster Radeon Pro460.
RAM/CPU performance / specs havnt changed much. Overall yes the system is faster / more responsive. I think the performance issue of the new MBP late2016 is the main issue with this MBP update but the reduction to USB-C, +the battery capacity issue – are just adding to this problem. Having 16GB RAM might seem to be enough for most but if you run a virtual machine often, than 32 GB RAM provides a real speed boost – and its NOT an upgrade option with the new MBP late2016 version. With full res smart previews in Adobe LR and working with the full res original files this MBP seems to be just overstrained, as was my late2013 version of the MBP. You have to shutdown everything else to let LR run full throttle and this is using the GPU allready. Maybe LR ist not yet able to use the GPU to its full potential, but I hardly see a big speed boost compared to the MBPlate2013 – and the late2013 was also a fully out-speced MBP.
Look at how these different MacBookPro generations (that I used since 2011) compare: the 2010 17inch is still running without hickups.
Fig.: Below: Mid2010 17 Inch MBP, middle: late2013 15 Inch MBP, above: late2016 15 Inch MBP, what is next? I will clearly have to add a hyperdrive.
Continue reading “The new MBP “late 2016” – a mixed experience …”
Using DroneDeploy and the POI mode of DJI (this article was first published on jenacopterlabs.de).
Have been doing some leaf-off flights now in March on one of my favorite sites for complex tree crown point cloud mapping. To add to the leaf-on data from last autumn.
Fig.: Leaf-Off point cloud with combined height color coding and reflectance color coding.
This time I also checked the full automatic flight modi using the POI (Point Of Interest) mode from DJI and Drone Deploy with the P4Pro and the P3A.
Both work perfectly although I believe the POI mode can be dangerous when its done with low altitudes. You have to carefully check that the radius is free from obstacles when you define the center position, altitude and distance from center (radius) for the POI flight. When POI is started it begins with comparably low speeds. You can modify the speed setting when the POI mode started. Unfortunately you cannot modify this stetting before you start. Tuning the speed on an iPad or (even more difficult – on an iPhone) is kind of dangerous because you easily move the slide too much to higher speeds than wanted and the copter will immediately accelerate and start circling your POI object like crazy until you managed to move the slider to slower speeds. There is however always this stop button to kill the POI mode – this is handy and needed sometimes. When you press the stop button the copter stops the POI mode completely and waits for new commands.
Continue reading “Drone Deploy and DJIs POI Mode copter image acquisitions”
Did some more tests with the Phantom 4 Pro:
The into the sun shots just capture a lot of flare and the sun star is very much Samyang 14mm like. This is an issue and it reduces the image acuity for 3D point cloud modeling for object measurements as well as for the landscape panorama approach.
The against the light shots look very much like the Samyang/Rokinon 14mm lens character. The sunstar has light rays that increase in size from the center and these rays cover the full frame of the 1inch sensor when the illumination comes from one of the outer edges. It clearly degrades the full image when the lens is stopped down to f.e. f8 but gets better when the lens is driven wide open. There seem to be some diffraction effects at work that are clearly not so great. You can drive the camera to generate sharp sun stars, but as usual here the characteristics of this star make a difference to some of us. The Samyang like star is not so well received. The linked flare issue however is much more a problem because it is hard to avoid when you have the sun in your frame and when the lens is stopped down to f8 or f11. For some scenarios a lens hood may be useful but its only functional for those shooting scenarios where the sun is not within your FoV.
To be honest I am bit puzzled … this is the first affordable 1-inch sensor platform with nice resolution and better DR, but – the flare issue is likely killing some of the potential ideas that you might develop with this machine. A pitty!
Comparison shot done with the Phantom 3A, clearly shows that the lens/sensor combi wont give the same amount of flare here (again different light level and different sun illumination angle):
Multicopter Photography and Photogrammetry
Late this year I bought a Phantom 3A (Advanced). The prices went down and these platforms are so affordable now that it makes sense to test it out.
The main applications for my work are multifolded:
- Special shootings for paying clients,
- regatta event shots from above,
- very early morning misty landscape stills and
- large scale or near range photogrammetric mapping for research purposes with applications in forest biomass mapping and archaeology as well as biotop-typ delineation and fine scale analysis of vegetation structure and slope stability analysis.
Clearly 1-3 and 4 will very nicely cooperate. Its just wonderful to experience how photography and research can come together here, makes me wonder sometimes how I managed to get into that cool spot where I get payed to fly a remote controlled camera … :-)
Being not new to copter flying, I found the Phantom 3 to be really easy to fly. Once you mastered the documentation and some basic theory about controlling a quadro-copter its really a simple excercise to control these devices. This is mainly due to the perfect GPS controlled position hold of the Phantom. It just stays where you put it (in opposite to gliders) and this makes it also kind of boring to fly a copter. The DJI GO app that is used on a mobile device to check the telemetry data and to control the lifeview provides a nice overview about position, viewing angle and overall status of your copter. You can also modify shooting parameters, change the camera viewing direction and check speed, height and remaining flight time, battery status and the number of satellites. Other apps support fully automatic mapping modes with variable overlap configurations (Data Mapper or Drone Deploy).
P3 installed on a LowePro backpack fully functional with installed props and ready to fly in a minute.
Some DJI Phantom 3 A Specs:
- Image Data capturing in 12MP Adobe DNG RAWs, gimbal stabilized! & damped. 2. HDR with 5 exposures each 0.7EV under/over-ex-> RAW DNG HDR computation. 3. Sony EXMOR 1/2.3“, 12.4MP, 94°lens (20mm focal length equiv. tilt.), ISO100-3200, 8-1/8000s, single shot, multiple shot, exposure series (3/5), DNG RAW format, video: FHD upto60p, 2.7k upto30p, 4. Memory: MicroSD up to 64GB cat10.
- Control: GPS/GLONASS controlled position hold, visual Realtime-Position-Tracking (VPS – Vision Positioning System) up to 3m flight alt. (P4pro: 10m). RTH (Return to Home Funktion), Battery Low RTH, Security-RTH, software controlled flight altitude (override possible): 120m, autostart/-landing, beginner modi, (OrbitModi/anti- collision (infrared/ultrasonic)/ActiveTracking/terrain follow: P4/pro/Mavic), P-GPS – GPS controled stabilising (+/-10cm position), P-OPTI (optical position control & stabilizing), P-ATTI modi (flight altitude only (barom.) – copter could start to drift).
- Flight parameters: weight: 1280g, climb: 5m/s, descend: 3m/s, Vmax: 16m/s (ATTI) – 60km/h, flight altitude max: 6000 m (airspace above 100 m needs a clearance (approval), RC max distance: 5km, flight time: 23 min (30min: P4pro), RadioControl: 2,4khz RC with iPad/iPhone (build in DJI Lightbridge System), realtime lifeview control per DJI GO App.
Continue reading “Multi Copter Photography and Photogrammetry”
Winter break for the Yacht/Regatta racing event photo shootings. Good to reorganize, sending some stuff for repair and to rethink about some approaches. I will also give some other sports a try this winter – some other water related sports … :)
Have been also looking into micro multicopter based photogrammetry with the Phantom lately – so there will be a small review about the capabilities here with more details soon.
Die letzte Galerie zur J70 German Open ist jetzt online. Durch die guten Bedingungen in Berlin bei diesem Event gibt es insgesamt recht viel ordentliches Bildmaterial. Mit Sponsorenveranstaltung/Eröffnung/Beamershow und Siegerehrung wars allerdings auch zeitlich eines der Events mit dem größeren Aufwand dieses Jahr.
Die 3te Galerie ist nun hier online: http://soeren.zenfolio.com/j70go2
Einige Ideen sind schlicht am Wetter gescheitert, die Über-Wasser Flüge mit der Drohne sind eine Herausforderung, da ein Teil des doppelten Netzes /RTH Funktion/ schlicht wegfällt auf dem Wasser, allerdings ist das aus der Luft fischen der Phantom vom Mobo aus erheblich einfacher als gedacht. Die Phantom fliegt auch bei schwachem Wind sehr stabil, Tests mit stärkerem Wind mache ich dann aber lieber auf einem großen Feld … . Durch den Formfaktor ist die neu herausgekommene Mavic Pro jedoch eindeutig eine interessante Wahl für diese Shootings. Bleibt zu testen, wie das mit der Leistung aussieht wenn Wind aufkommt. Die Bildqualität dieser kleinen Sensoren ist grenzwertig und eignet sich (auch wenn man alle Register zieht – HDR/LR Nachbearbeitung / Pano Mosaikierung – nur für Plots bis etwas über A3. Das dürfte mit der Mavic Pro kaum besser geworden sein. Mittelfristig muss eine APS-C Kamera an die P3. Hier in der engeren Auswahl ist eine alte Cam von Rico und die A6300 von Sony, ggf auch eine Vollformat A7. Wenn man da ankommt, dann ist es fix auch sinnvoller an den Einsatz einer Inspire 1 Pro mit ZM X5 zu denken.
Am letzten Tag der J70 German Open gab es dann nochmal sehr wechselhafte Bedingungen mit mehreren Schauerzellen, die recht schöne Lichtmischungen mit der wieder durchbrechenden Sonne erzeugten. Nur für die Drohne war der Tag recht schlecht geeignet. Als dann endlich dauerhafter die Sonne durchkam, war die letzte Wettfahrt auch gelaufen. Durchaus ein Problem ist, das diese Drohnen einfach gar keinen Regen vertragen. Man kann daher bei etwas wechselhaften Bedingungen im Prinzip einpacken und sich auf etwas anderes konzentrieren. Hier ein Shot von der J70 German Open. Um das Maximum an Bildqualität aus dem bescheiden großen Sensor der Phantom 3 zu holen, muss man schlicht HDR Aufnahmen machen (ohne auch den klassischen HDR Effekt erzeugen zu wollen). Da die Drohne recht beeindruckend gut ihre Position hält und der Gimbal die Ausrichtung der Kamera korrigiert sind Mehrfachbelichtungen ganz gut möglich. Dieser Shot wurde allerdings erst im Post kombiniert. Das eröffnet einem eine ganze Reihe von Möglichkeiten bei der Visualisierung von Bewegungen. Bin mir noch nicht so sicher was alles letztendlich realisierbar ist. Im Prinzip muss man sich ja über dem Regattfeld positionieren und das ist überhaupt nicht vereinbar mit den Aufstiegsbestimmungen von Drohnen. mtk, Sören
J70 German Open 2016, Gr. Fenster/Havel/Berlin.
Update: some nice bokeh shots with this lens (had half an hour yesterday with wonderful sun light to do some tests – its easy to send every for/background into a creamy bokeh fog with a 400mm f2.8 system. However the lens isnt a dedicated macro device – hence it will be important to see how the lens performs in the mid to far range shots and with full reflecting highlights. Especially the resolving power with very distant objects is usually a problem – even with these top-of-the-line tele lenses. My first impressions indicate that the 400mm f2.8 works fabulous with the TC14E II and also very good with the TC20E III but the AFS 300mm f2.8 mk1 wasn’t bad with the TC14E II either. The weak part was the resolving power focusing more distant things with more than 300m distance.
First tests with the new 400mm lens – the f2.8 AFS VR version – pretty much a dream tele lens and optical performance is just as what you would expect. First lens align tests seem to be spot on. The lens snaps to focus – clearly faster than my old 300mm f2.8 AFS Mk1. Some bokeh tests will follow.
The lensAlign target with the long ruler installed – approx. 20m distance – this equals 50times focal length. At f2.8 this gives a depth of field of 40cm and the lens is pretty spot-on on the 0-Mark of the ruler. No chromatic aberration at all … . Damn cool lens – only point to think of right now is the how-to-story of “hand-holding” this lens, its just bigger than anything I used before.
Some bokeh shots: